On Sunday, May 26th during 2:00AM-5:00AM we will be upgrading our core
routing and switching capacity. I have dissected the upgrade path that
failed last time, and I believe I know which Cisco bugs I was running
up against when patching the new core switch stack into our network.
I was able to downrev my test rig to the exact same version with
simular results, and have made sure to have a method that won't
trigger those bugs, with several alternate upgrade path options.
We will be swapping out older gear (Juniper M20s that have worked
extremely well for years and years) with new current models (Juniper
MX80) with much greater capacity. Our core switching network will also
be upgraded to dual 10G aggregates at this time, although the access
layer to the rest of the network will continue to be the same gear we
have been using.
There will also be a a 10G upstream feed brought in to supplement
our other Internet upstream connections. This will be coming
in from CenturyLink. An upstream we've used in the past, so we are
familure with them. This will give us plenty of extra capacity as
needed from our upstream feeds.
We expect the disruption to be minimal, although there will be some
downtime while routing is being cut over. The new gear has been
burning in for some time, while we were becoming familure with the new
quirks that new gear always brings, as well as finding the bugs and
gotchas in getting a design that is workable for our needs.
This will affect just about every service in some fashion due to the
extensive cutover, which is why it is being done in the lowest traffic
time period. We will be continue monitoring network wide to ensure all
services are returned to normal after the cutover.
If you have any problems or questions please let us know at
[log in to unmask], or call us up at 612-337-6340.
--
Doug McIntyre <[log in to unmask]>
-- ipHouse/Goldengate/Bitstream/ProNS --
Network Engineer/Provisioning/Jack of all Trades
